Statistics from Altmetric.com
Successful treatment of CMV ventriculitis immune reconstitution syndrome
In HIV infected patients, cytomegalovirus (CMV) disease of the central nervous system is usually seen when the CD4+ count is less than 100 cells/mm3. Recommended treatment includes intravenous ganciclovir, foscarnet, or both.1 Ventriculitis is an uncommon but severe manifestation of CMV disease. Mortality is extremely high despite timely initiation of treatment. Little is known about CMV ventriculoencephalitis in the highly active antiretroviral therapy (HAART) era.
A 38 year old woman with no significant past medical history was diagnosed with Pneumocystis jiroveci (Pneumocystis carinii) pneumonia as her AIDS defining illness in April 2002. On 15 July 2002, HAART was initiated with zidovudine, lamivudine, and efavirenz. Her CD4+ count was 42 cells/mm3 and her HIV-1 RNA level was 95 000 copies/ml. Serum CMV IgG was positive.
Ten days later, she was evaluated for headache of new onset. Computed tomography (CT) of her head was unremarkable. Cerebrospinal fluid analysis showed no white cells, a red cell count of 10/mm3, a protein level of 58 mg/dl, and a glucose of 40 mg/ml. Cryptococcal antigen serology was negative.
Two weeks later, after full compliance with her HIV therapy, she returned with headache, nausea, vomiting, and fever. Her temperature was 38.6°C (101.4°F). Neurological examination was normal and there was no meningism. Ophthalmological examination of the retinae with fully dilated pupils was normal.
The peripheral white blood count was 1800 cells/mm3. Serum chemistry was normal. Cryptococcal and histoplasma antigens were negative. Bacterial, mycobacterial, and fungal blood cultures were sterile. Chest radiography and repeat head CT were normal. CSF analysis showed 13 white blood cells/mm3 (44% neutrophils, 44% lymphocytes, 12% monocytes), a red cell count of 4 cells/mm3, a protein level of 91 mg/dl, and a glucose of 30 mg/ml. Magnetic resonance imaging (MRI) of the brain showed abnormal contrast enhancement of the ependymal lining around the ventricles, consistent with ventriculitis (fig 1). CSF herpes simplex virus polymerase chain reaction (PCR) was negative. Both CSF PCR and bone marrow culture were positive for CMV.
Ganciclovir (5 mg/kg intravenously twice daily) was initiated. Seven days later, the patient was afebrile, but noted only mild improvement in her headache, nausea, and vomiting. Therefore, intravenous foscarnet (90 mg/kg twice daily) was added; her symptoms resolved within 21 days of combined therapy. At that time, her CD4+ count was 159 cells/mm3 and the HIV-1 RNA level was <400 copies/ml. Follow up MRI showed improved appearance of the enhancement that was present at her initial diagnosis. CSF CMV PCR was subsequently negative. Valganciclovir (900 mg orally daily)2 for maintenance treatment was added to her HAART regimen. Three months after completing CMV induction therapy, the patient’s CD4+ count was 178 cells/mm3 and HIV-1 RNA was <50 copies/ml. Valganciclovir maintenance therapy was discontinued. There has been no recurrence of CMV disease over the 16 month period since discontinuing valganciclovir.
Several points in this case are noteworthy. The patient was diagnosed as having CMV ventriculitis three weeks after initiating HAART, during a period of immune recovery. While immune reconstitution syndromes (IRS) involving CMV retinitis and pneumonitis have been reported, to our knowledge an IRS with CMV ventriculitis has not been described previously. The time to onset of IRS associated with CMV infections ranges from weeks to several months. Our patient developed a headache with a normal laboratory evaluation after 10 days of treatment. Three weeks after initiating HAART, her illness progressed to include fevers, persistent headache, nausea, and vomiting. While we cannot absolutely exclude the possibility that her CMV ventriculitis was a coincidental infection of a susceptible individual, we believe the timing of her illness with recent initiation of HAART is more consistent with a CMV associated IRS. Our patient was CMV seropositive and therefore latently infected when HAART was initiated. Her CD4+ count increased from 43 to 159 cells/mm3 after 45 days of HAART. Additionally, between the 10th and 31st day of HAART, she developed CSF leucocytosis with progressive CSF protein elevation providing evidence of a significant CNS inflammatory reaction in a time interval consistent with IRS.
Studies addressing the best treatment for CMV ventriculoencephalitis in patients with AIDS are not definitive, although various comparative trials have addressed the treatment of retinitis. Current treatment options for CMV neurological disease include intravenous ganciclovir, intravenous foscarnet, or both.1 In a study of HIV infected patients with acute CMV infection of the CNS, combination foscarnet-ganciclovir therapy resulted in clinical improvement or stabilisation in 23 of 31 patients, but 10 of these had significant side effects necessitating discontinuation of at least one drug.3 Our patient tolerated both foscarnet and ganciclovir for 21 days without significant side effects.
Given her initial presentation with advanced HIV-1 infection, we placed the patient on valganciclovir maintenance therapy (900 mg once daily) to prevent relapse of her CMV ventriculitis. In previous reports, maintenance therapy with intravenous ganciclovir prevented relapse of CMV encephalitis or myelitis in 13 of 23 AIDS patients.4 Linear regression analysis of data from CMV and HIV seropositive patients shows that valganciclovir, 900 mg once daily, produces target AUC24 values comparable to those achieved with intravenous ganciclovir at a dose of 5 mg/kg/day.5 Our patient with CMV ventriculitis was prescribed oral valganciclovir maintenance therapy. Discontinuing maintenance therapy for CMV retinitis in patients who responded to HAART has been successful.5 It is reasonable to follow the USPHS/IDSA guidelines for CMV retinitis and discontinue maintenance therapy when there has been a sustained (for example, more than six months) increase in CD4+ counts to >100–150 cells/µl in response to HAART. Based on our patient’s robust CD4+ count improvement and complete suppression of HIV-1 viraemia, we judged her risk for CMV disease to be low and discontinued valganciclovir when her CD4+ count had been >100 cells/µl for three months. She has had no evidence of relapse of her CMV disease two years later.
In conclusion, CMV ventriculitis can present as an IRS in AIDS patients and should be considered in the differential diagnosis in a patient with CNS complaints, even in the absence of retinal disease. Aggressive treatment should be pursued, given the high mortality of the disease, and combined ganciclovir-foscarnet may be indicated based on the patient’s presentation. The use of oral valganciclovir is a reasonable option for maintenance therapy.
Competing interests: none declared
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.